From what I understand some colleges initially reduce their application pool using minimum GPA and ACT/SAT scores. How common is this? Any AO’s out there with real world knowledge?
I have seen some AOs suggest that they may do an academic review first, and for those unhooked applicants that do not meet their normal academic standards, those applicants may only get a quick “holistic” look to see if something stands out to make an exception. If not, they may (in some form) be moved relatively quickly to a presumptive denial pile.
1 Like
Look at the bottom 25% threshold for GPA/test score: students in that group have something else of interest to the college. If you don’t have anything special and are within that tier stats-wise, expect to be cut. Same thing for bottom 50% but less strictly so.
And of course selectivity matters: a college with a sub 20% acceptance rate will expect “exceptional” everything within the context of your HS/community, whereas a college with a 75% acceptance rate will admit you provided you meet their basic academic standards (rigor×GPA) ECs and other achievements are cherry on the cake.
4 Likes
This was discussed a couple of months ago on the podcast Your College Bound Kid. Apparently, Yale is now doing a prescreen because they get too many applications that take up too much resource. The test optional era is truly here. IMO the popular schools will implement some for of this.
3 Likes
Admission Departments have access to software which collates unhooked applications and essentially applies cuts which instantly eliminate a significant batch of applicants. some schools will assign one lucky AO to have a cursery look at that “eliminated” batch to confirm it with a pair of human eyes. Been told that that insta reject batch can reach 50% of the applicant pool in one instance, and can only assume many schools use the same software, obviously sitting there program differently. this is the natural result of schools which have consistently kept marketing themselves to reduce their admisdion rate significantly over the years, at a much faster rate than actual population growth
3 Likes
Do you have a citation to support this and the rest of your post?
1 Like
confidential unnamed source. pretty sure no school will openly disclose what software system they use and how, but more specific information will inevitably be brought in the open with time.
4 Likes
So…no is the answer. Maybe let’s stick to opinions - sans debate - since I doubt any AO is sharing this information publicly.
6 Likes
Dartmouth’s Dean of Admission addressed usage of AI in reviewing applications here (toward bottom - search for word preliminary):
https://admissions.dartmouth.edu/follow/admissions-beat-podcast/admissions-beat-s3e15-transcript
This article might also be of interest and discusses automated transcript review:
1 Like
Almost all colleges that practice holistic admissions increasingly need help screening their applicants. I have no doubt that many more of them are experimenting with the use of AI tools and their use will expand from initial screening into many areas of college admissions over time.
I agree they are experimenting with AI, with that said many of the schools we are talking about use external readers, and a lot of them.
For example, UCLA, the most applied to college, uses 200-250 external readers each year. I don’t know of any holistic highly rejective school that does not read each app fully at least once. UCLA commits to reading each app twice, in its entirety.
1 Like
This is a great video that actually goes through in real time (a bit sanitized but still awesome) how a file is read at a selective school (College of Holy Cross).
It was shared as a resource on YCBK podcast (episode 349) and I am just the messenger/copy paste guy.
Most colleges use a standard software for preparing files (SLATE). This is well known. Slate allows for customization. Yale is doing cutoffs and I just heard on the same podcast that based on U Chicago rep, approximately 65% of their applicants now are academically qualified (thanks to test optional). @mynameiswhatever @NiVo
6 Likes
@1dadinNC this is a great visual of the process, and one can predict how SLATE will empower their software to accelerate many steps.
your U Chicago comment would essentially mean the software can eliminate 35% of the applicant pool (or alternatively push it to an accelerated review to confirm it’s not suitable), so AOs can more efficiently focus their time on the 65% qualified applicants.
one can imagine that AI progress would only increase that “automated” 35% percentage over time. indeed, time will tell.
3 Likes
Slate has been in use to read apps for years. UChicago commits to read every app in a holistic manner. If some holistic schools change their process dramatically to utilize AI, I hope they are transparent about that. Chicago would have to change their website/FAQs:
UChicago welcomes students from all backgrounds, and each application at UChicago goes through the same holistic review process.
What is your minimum GPA or required SAT or ACT score? There is no minimum GPA or required test score. At UChicago, the admissions committee considers a candidate’s entire application—academic and extracurricular records, essays, letters of recommendation, and optional testing—and there’s no one piece of information that alone determines whether or not you would be a good fit for the College. You can learn more about this holistic review process here.
Is there a score cut off at which I should opt out of submitting my ACT or SAT? We review applications holistically, which means there is never a score “cutoff” that would determine the fate of a student’s application.
Aside from potentially needing to hire fewer admissions officers and/or external readers (which are relatively cheap labor), why would a college that currently uses holistic admissions want to stop doing so?
1 Like
As I understand it, screening holistic review colleges typically still do a quick look at the whole application before putting it in the presumptive reject pile. And they may still have two reviewers do that. If something stands out as worth more consideration, they will then give the whole application a longer review. This doesn’t happen often, but it does happen sometimes.
I think in their view, that still counts as holistic review, and is consistent with saying there are no hard cutoffs, and so on. But it will go quickly if you don’t meet their normal academic standards and then nothing unusual stands out.
That’s consistent with what I wrote. Average app review time is 6-10 mins, plenty of time to get thru the whole app. At some schools apps will go to committee after the first read or two as well.
1 Like
By the way, it appears to me that Dartmouth transcript linked contains a good example of that sort of logic. In the relevant section, Dean Coffin first says this:
I have stubbornly held to the idea that we need to read them all the way through holistically to learn their story, and my colleagues will tell you, he does make us do that and write it out in evaluative form, but it’s pushed against the limit of our ability to do that, given the volume, the date of the calendar.
But he then also says this:
[W]e are using AI to a degree, so, guilty. We are starting to sort the preliminary pool in some ways, because we have to figure out, what’s the first wave? What’s the second wave? Which students are structurally more challenged by their academic elements, and they move into a faster track . . . .
Again I think at least the way they view things at Dartmouth, using AI and similar tools to identify applicants who “are structurally more challenged by their academic elements” and then putting those applicants on “a faster track” is still consistent with “the idea that we need to read them all the way through holistically to learn their story.” Because it is still a form of holistic review, just a very quick form in most such cases.
AI for sorting makes sense.
No different than the schools that decile apps based on each applicant’s individual predicted yield, as calculated by the school’s predictive analytic model. The apps in the higher predicted yield deciles get a more detailed read, but all of the apps across all deciles are still read in their entirety. AFAIK.
So I just quoted Dartmouth’s Dean Coffin as suggesting AI is part of the process of putting some academically “challenged” applicants into a “faster” track (presumably that is normally a faster track to a reject decision).
This implies to me that for those applications, they are looking to get to a presumptive decision faster than average, to leave more time for other applications.
I agree that from their perspective, this is still a form of holistic review. But, it is also a form of using AI and such to prescreen applicants. It combines elements of both in an effort to more efficiently allocate review time while still being holistic.